Guest Presentations

Tina Tallon, Assistant Professor of AI and Music Composition, The Ohio State University

From the Turing test to the recent proliferation of large language models (LLMs), post-war developments in generative AI have largely focused on language as a proxy and evaluative mechanism for human intelligence. Conversely, musicians have long noted the inadequacy of language to drive and describe our creative practices. This highlights a friction between the use of LLMs as means of interfacing with generative music tools and the non-linguistic specificity and embodied knowledge involved in music-making. Textual interfaces are becoming increasingly common as a means of interacting with generative music systems, most notably in the form of text-to-music (TTM) models. Companies developing these tools tout the increased access that textual interfaces will provide, but many musicians still find TTM models unable to provide the detailed control that they need to create their work. Additionally, these tools raise concerns about everything from the devaluation of foundational skills involved in music-making to the exploitive ways in which copyrighted material has been used to train these models without artists’ consent. This paper audits five recent publicly-available datasets used for the training and evaluation of TTM models, examining data provenance, correlation with common features obtained through music information retrieval (MIR), and prevalence of bias. While some of the TTM datasets admirably take steps to increase utility and mitigate potential harms, many still exhibit an array of limitations related to opacity of data provenance, prevalence of blatant inaccuracies in data labeling, and overt intersectional biases. While TTM models may in fact have the potential to expand access to and efficiency of some tools used in music-making, they should be developed in close collaboration with a wide array of musicians who work in a diversity of genres, aesthetics, and traditions in order to ensure that they are truly tools of co-creation and not exploitation.

Winner of the 2022 Rome Prize in Music Composition, Tina Tallon is creative technologist and composer whose work explores the myriad ways in which AI impacts how artists engage with society. Her music, interactive installations, and research has been presented globally in diverse settings, including Disney Hall, the Large Hadron Collider, major motion pictures, and leading AI conferences like NeurIPS. Tallon has earned numerous awards and grants from organizations such as Harvard, MIT, and the American Academy of Arts and Letters, and she currently serves as Assistant Professor of AI and Music Composition at The Ohio State University.

Teerath Majumder, Assistant Professor of Sound Design, Columbia College Chicago

As our lives become more and more intertwined with so-called artificial intelligence (AI), we are challenged with difficult questions about the future of human-machine relationships. While many believed that AI would be used to solve some of the most prescient issues of our time (such as detecting cancer at early stages), we are witnessing profit-driven industries replacing their human workforce with AI. Those in creative fields are not immune to this change; the proliferation of AI text/image/audio/video generators have already affected creatives in many areas. In the presence of such evidence, it would be naive to assume that corporations investing heavily in generative AI are not doing so to bypass the need for human creators.

With this reality in mind, my collaborator James Ilgenfritz and I have been investigating what a sustainable human-machine relationship may look like, one that is not concerned with efficiency and profit but with understanding and collaboration. To that end, we have been developing the project Dysnomia where we create and interact with complex systems of physical and computational agents, often adopting recursive and feedback processes. Instead of applying pre-trained machine learning (ML) models, we purposefully design our own models and train them with our own data (often in realtime) to gain a deep understanding of their intricacies. We believe this understanding is key to transcending our exploitative tendencies and building a healthy human-machine relationship.

Teerath Majumder is a Bangladeshi composer and technologist who works in interactive and immersive media, computer music, and sound design. He questions socio-sonic dynamics that are often taken for granted, and reimagines relationships between participants through technological mediation. Notably, he produced Space Within (2022) where audience members collaborated with featured musicians to give rise to an hour-long sonic experience. His current project titled Dysnomia is an investigation of machine agency in mediating human relationships where he “collaborates” with artificial intelligence. Teerath is an Assistant Professor of Sound Design at Columbia College Chicago.

Anna Heflin, DMA Candidate in Composition, University of Southern California

This presentation will examine the decades-long collaboration between the Advancement of Creative Musicians (AACM) and Czech-born composer Petr Kotik’s New York-based S.E.M. Ensemble and Ostrava Days Festival in the Czech Republic. Petr Kotik and The Orchestra of the S.E.M. Ensemble began collaborating with AACM composers in 1995, when Muhal Richard Abrams invited Kotik and his ensemble to work together on a series of AACM orchestra concerts at the New York Society for Ethical Culture. Between 1995 and 1997, Kotik conducted orchestra compositions by Amina Claudine Myers, Roscoe Mitchell, George Lewis, Wadada Leo Smith, Joseph Jarman, Henry Threadgill, and Muhal R. Abrams. Kotik founded Ostrava Days in 2001, resulting in an international component to the collaboration. The musical focus of my presentation will be on the large-scale works performed in Ostrava by AACM composers, including the 2015 premiere of George Lewis’ Afterword from his opera A Power Stronger Than Itself, the 2023 performance of Amina Claudine Myers’ Improvisational Suite (1979), and Roscoe Mitchell’s 2020 recording with Ostravská Banda. Research methods will include score study, interviews with the artists (contingent on their permission), archival recordings and interviews from Ostrava Days, and existing published research. S.E.M. Ensemble has celebrated this collaboration in the form of concerts, namely with concerts in 2015 and 2023. While existing research abounds about AACM, I have not found published papers with this emphasis on the social and musical relations between AACM and Ostrava Days.

Anna Heflin is a composer and writer who constructs high-octane, humorous, and sensual worlds with non-linear narratives that thrive on musical and psychological fragmentation. Whether writing a symphony or a staged literature-inspired solo opera for an instrumentalist, she is drawn to the unexpected and channels her highly imaginative virtuosic visions into complex characters and unorthodox narrative arcs that often integrate text and staging. Her long-term collaborations with individual artists and organizations developed over years of working as a freelance violist are central to her process and her core values include trust, risk taking, experimentation, play, open communication, and creative problem solving.

Sasha Ishov, DMA, University of California San Diego

How does technology reshape performer-composer communication and foster shared authorship? This presentation explores co-composition and creative authority from the performer’s perspective, based on my dissertation "PrismaSonus: Bridging Acoustic and Digital Worlds in Flute Practice and Performer-Composer Collaboration," where I write about a project developed with composer Theocharis Papatrechas while I was a student at UC San Diego. 

The project investigates the relationship between microphone placement, flute technique, and perception, integrating practical experimentation with theoretical frameworks from human-computer interaction, cognitive science, and phenomenology. Placing microphones inside the flute unveils hidden timbres and microtechniques, displacing traditional listening perspectives and revealing new sonic landscapes. Iterative cycles of recording, listening, and processing synthesized technique and technology, cultivating a shared understanding of instrumental affordances. 

This dialogue resulted in three formats—fixed media, live performance, and an online experience—each engaging listeners with the flute’s “inner world.” These works highlight how technology, notation, and technique can challenge traditional roles and foster shared creative language, artifacts, and knowledge, transcending conventional hierarchies.

The presentation addresses three themes: 

1.Embodied Perspectives: Technology alters performers’ sensory and auditory relationships with instruments, requiring adaptation to unfamiliar feedback.

2. Documentation as Transformation: Iterative recording and listening processes shaped artistic outcomes and co-creative dynamics.

3. Blurred Boundaries: Collaboration disrupted traditional performer-composer roles, fostering new approaches to creativity and co-composition.

PrismaSonus offers a model for how technological interventions can serve as active agents in redefining musical roles, broadening the expressive potential of instruments, and fostering innovative collaborations.

Our project highlights how performer-informed electroacoustic experimentation can create new alliances and reimagine creative authorship. By embracing technology, the project offers a model for performer-composer partnerships that challenge and expand traditional boundaries.

Sasha Ishov is a flutist, educator, and researcher redefining the flute’s role in contemporary music. Praised for his “well-sounded and lucid” artistry (San Diego Union-Tribune), he has performed at the Ojai Music Festival, BBC Proms, Carnegie Hall, and with the Aspen Contemporary Ensemble. A champion of new music, he has premiered over 100 works and collaborated with composers including John Luther Adams, Augusta Read Thomas, and Matthew Aucoin. He co-leads Offscreen, a duo with percussionist Michael Jones, and researches how technology shapes collaboration. Sasha holds a DMA from UC San Diego, a BM from Eastman, and is a Miyazawa Artist.